Datasets | Mozilla Data Collective

All Datasets

Compensated Datasets

API

Contribute Datasets

Loading datasets...

Terms

FAQs

Participation Guidelines

Responsible Disclosure

News and Documentation

Data and Model Competitions

[email protected]

Terms

FAQs

Participation Guidelines

Responsible Disclosure

News and Documentation

Data and Model Competitions

[email protected]

Mozilla Data Collective is a mission-locked British company, incubated by Mozilla Foundation, and backed by the not-for-profit Mozilla.org

All Datasets

Not sure where to start?

Describe what you are building and the Data Assistant will find the right datasets for you.

Try the Data Assistant

Filtered by

Search results for “Dari 1 Million Text Corpus”

100 Datasets

MirasAI

Associated Press Pakistan 10.4 Million Saraiki Corpus

A 10.4-million-word Saraiki news corpus from Associated Press of Pakistan, covering 2016–2026 for non-commercial Saraiki NLP research.

License: CC-BY-NC-SA-4.0

Locale: skr

Task: NLP

Format: TXT

Size: 19.51 MB

TAUSCompensated

🚧 Colloquial Parallel Text (English - Yoruba) 🚧

49k native-reviewed colloquial sentences translated from US English into Yoruba for fine-tuning conversational MT and AI models.

License: TAUS DUA

Locale: yo-NG, en-US

Task: MT

Format: TSV

Size: 3.25 MB

Kaleem Art Press

Kaleem Art Press Saraiki Literature Corpus

This corpus contains approximately one million tokens of Saraiki text curated over the past ten years by Kaleem Art Press. It features a wide range of literary genres, including stories, short stories, novels, fiction, non-fiction, travelogues, poetry, biographies, and historical writings. All content is shared with full author approval. The dataset is intended to support linguistic research, Saraiki language technology development, and the preservation of cultural and literary heritage.

License: CC-BY-NC-4.0

Locale: skr

Task: OTH

Format: TXT

Size: 1.84 MB

CLEAR Global

TWB Parallel Sentence kits - Kanuri (5k)

5,000 English–Kanuri parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

License: CC-BY-4.0

Locale: kau

Task: MT

Format: TSV

Size: 358.46 KB

Digital Divide Data

Khmer ASR Cultural Dataset (V2)

106.53 hours manually curated speech-text pairs by native speakers in Khmer language about Cambodian cultural topics. On average, each recording is 8.42 seconds with the standard deviation of 3.39. Speaker metadata (gender, age group, and origin city) is provided. - Language: Khmer (khm). - Source(s): Native speakers from Cambodia (4 females, 4 males). The utterances were manually generated based on topics and subtopics listed in metadata. - Domain(s): Cultural domain, with a total of 2 topics and 118 subtopics. - Size: 45.57k data instances - WAV file names are formatted as: `{speaker_id}_khm_{sentence_id}.wav`.

License: CC-BY-SA-4.0

Locale: khm

Task: ASR

Format: WAV

Size: 35.86 GB

Habib University

Burushaski-English Speech Translation Corpus

A linguistically informed Burushaski speech - English text corpus containing 14,970 short audio recordings paired with corresponding English text translations.

License: CC-BY-NC-SA-4.0

Locale: bsk

Task: MT

Format: WAV, TXT

Size: 921.47 MB

Protham Alo

Prothom Alo Bengali Newspaper Corpus

A 10+ million token Bengali newspaper corpus from Prothom Alo provided in .docx format for large-scale NLP research and linguistic analysis.

License: CC-BY-NC-4.0

Locale: ben

Task: NLP

Format: DOCX

Size: 42.36 MB

Taruen

Dolgan Folklore Text Corpus

A 15.6k-word corpus of 19 Dolgan fairy tales, digitized to catalyze NLP research and revitalization for this highly endangered Siberian Turkic language.

License: CC0-1.0

Locale: dlg

Task: NLP

Format: TXT

Size: 57.16 KB

Open Home Foundation

Imre 1.0

Text to speech dataset for Hungarian, male speaker, approximately 1.5 hours of read speech.

License: CC0-1.0

Locale: hu-HU

Task: TTS

Format: WEBM

Size: 99.60 MB

Community

Betawi TTS of Cultural Language (BEKAL)

Betawi TTS of Cultural Language (BEKAL) this dataset uses the Betawi dialect of West Java with Indonesian code-mixing and code-switching.

License: CC-BY-SA-4.0

Locale: bew

Task: TTS

Format: WEBM, TSV

Size: 309.99 MB

Habib University

Burushaski-English Speech Translation Corpus

A linguistically informed Burushaski speech - English text corpus containing 14,970 short audio recordings paired with corresponding English text translations.

License: CC-BY-NC-SA-4.0

Locale: bsk

Task: MT

Format: WAV, TXT

Size: 921.47 MB

Community

Indonesian Lip-Record Multimodal Corpus

This corpus comprises video footage of mouth movements, audio, and text transcriptions for 6 hours.

License: CC-BY-SA-4.0

Locale: ind

Task: OTH

Format: WEBM, TSV

Size: 984.41 MB

Taruen

Tatar Folklore Text Corpus

A 485k-word Tatar folklore corpus from 20th-century field recordings, selected from 5 academic volumes to prioritize contemporary linguistic usage.

License: CC0-1.0

Locale: tt

Task: NLP

Format: TXT

Size: 1.40 MB

Institute of Finno-Ugric/Uralic Studies, University of Hamburg

INEL Kamas Speech Corpus

A 14-hour Speech-to-Text dataset for Kamas, an extinct Samoyedic language waiting to be revitalized. Features audio aligned with text transcriptions.

License: CC-BY-NC-SA-4.0

Locale: xas

Task: ASR

Format: TSV, MP3

Size: 376.64 MB

Bahasa Bahana

BAHANA-Speech Corpus of English Learners from Indonesia

A corpus of read speech by learners of English from Indonesia.

License: CC-BY-SA-4.0

Locale: en

Task: CALL

Format: WEBM, WAV, TSV

Size: 1.14 GB

Institute of Finno-Ugric/Uralic Studies, University of Hamburg

INEL Enets Speech Corpus

A 3.7-hour supervised Speech-to-Text dataset for Enets, an endangered Samoyedic language. Features sentence-level audio aligned with text transcriptions.

License: CC-BY-NC-SA-4.0

Locale: enf, enh

Task: ASR

Format: TSV, MP3

Size: 140.56 MB

Kaler Kontho

Kaler Kantho Bengali Newspaper Corpus

A 10+ million token Bengali newspaper corpus from Kaler Kantho provided in .docx format for large-scale NLP research and linguistic analysis.

License: CC-BY-NC-4.0

Locale: ben

Task: NLP

Format: DOCX

Size: 33.11 MB

TAUSCompensated

🚧 Colloquial Parallel Text (English - Hausa) 🚧

49k native-reviewed colloquial sentences translated from US English into Hausa for fine-tuning conversational MT and AI models.

License: TAUS DUA

Locale: en-US, ha-NG

Task: MT

Format: TSV

Size: 2.93 MB

Open Home Foundation

Mihai 1.0

Text to speech dataset for Romanian, male speaker, approximately 2 hours of read speech.

License: CC0-1.0

Locale: ro-RO

Task: TTS

Format: WEBM

Size: 66.31 MB

Community

Sindhi Corpus

SdCorpus is a large scale monolingual Sindhi text corpus collected from publicly available web sources for NLP and language model development.

License: CC-BY-NC-4.0

Locale: snd

Task: LM

Format: TXT

Size: 321.00 MB

Community

Bojonegoro Javanese TTS

This dataset contains synthetic speech data covering a variety of everyday-life topics in the Bojonegoro dialect of Javanese, spoken in East Java, Indonesia.

License: CC-BY-SA-4.0

Locale: jav

Task: TTS

Format: .tar.gz, WEBM

Size: 469.50 MB

MirasAI

Noakhalian (নোয়াখাইল্লা) Text Corpus

A 500K token corpus of Noakhalian linguistic data including drama scripts and cultural texts for professional NLP research and language preservation.

License: CC-BY-NC-4.0

Locale: oak

Task: NLP

Format: TXT, DOCX

Size: 3.61 MB

Open Home Foundation

Tugão 1.0

Text to speech dataset for Portuguese, male speaker, approximately 1.5 hours of read speech.

License: CC0-1.0

Locale: pt-PT

Task: TTS

Format: WEBM

Size: 61.84 MB

Community

Gresik Dialect of Javanese-IND-ENG-ARB Parallel Corpus

This dataset comprises a parallel corpus of 5,173 lines of literal translation text in Indonesian, the Gresik dialect of Javanese, English, and Arabic.

License: CC-BY-NC-SA-4.0

Locale: ind,jav,eng,arb

Task: ML

Format: TSV

Size: 550.10 KB